BIOLOGICAL MASS SPECTROMETRY, VOL. 20, 515-521 (1991)
The Matrix Effect in Particle Beam Liquid Chromatography/MassSpectrometry and Reliable Quantification by Isotope Dilution F. Reber Brown and William M. Draper? Hazardous Materials Laboratory, 2151 Berkeley Way, Berkeley, California 94704, USA
The transport efficiency of the padcle beam liquid chromatography/mnss spectrometer interface is idueaeed by a d y t e concentration contributing to a widely reported non-linearity. In this work, coeluting, isotopelabeled interMIstandards were investigated as 'carriers' to improve the transport efficiency and linearity. Three styrene metabolitemadelk, phenyldyoxylic and bippuric acids-and tbei pentadeutero analogs were separated by reversed-phase liquid chromatogrophy (LC) with an ammonium acetate-acetonitrile mobile phase. Selected positive ions produd by electron ionization were monitored to generate pfllicle beam LC/MS calibration curves. The present study demonstrates that particle beam LC/MS not only is ma-linear, but also is subject to a matrix e&ect presumably by the same mechanism responsible for non-linearity. Caeluting, isotopelabeled internal standards were ineffective at linearizing the pnrticle beam liquid chromatograph/mass spectrometer detector response. Isotope dilution quantification, however, compensates for variable transport efficiencies, linearizes calibration and compensates for the matrix e&ect, affording reliable quantification of the styrene metabolites.
INTRODUCTION Particle beam liquid chromatography/mass spectrometry (PB LC/MS) has become popular because of ease of operation, compatibility with many LC mobile phases and the ability to operate in both electron (EI) and chemical ionization (CI) modes. Generation of library-searchable, EI spectra by PB LC/MS is an important advantage over other LC/MS techniques, and this capability is critical in regulatory monitoring and identification of unknowns. By comparison, direct liquid introduction LC/MS generates only CI spectra,'-' and thermospray LC/MS provides principally molecular weight inf~rmation.~.' Finally, because PB LC/MS uses a conventional mass spectrometric source, particle beam instruments are compatible with other mass spectrometric techniques such as GC/MS and sample introduction by solid probe. The development of PB LC/MS was first described by Willoughby and Browner in 19844 and the technique has been r e v i e ~ e d . ~As . ~ .PB ~ LC/MS is less than ten years old, and commercial PB interfaces have only been available for a few years, many of the basic features of the technique are only now being investigated. Recent reports have described characteristics of tuning, mobile phase effects, signal optimization and so forth.'-' Quantitative PB LC/MS has been hampered by nonlinear detector response. PB LC/MS non-linearity has been observed in various laboratoriesa*'0-'2 although at this time it has not been studied thoroughly or sys-
tematically. PB LC/MS calibration curves have been reported to be quadratic or second-order,a*'Obased on regression analysis of limited data sets. Trial and error and serendipity have led to means for minimizing and compensating for PB LC/MS non-linearity. Kim et al. added malic acid to the mobile phase to aid in linearizing the PB LC/MS response of daminozide." Similarly, Bellar and coworkers observed that ammonium acetate as a mobile phase modifier acted as a 'carrier' and extended the PB LC/MS linear dynamic range.I2 Isotope dilution (ID) mass spectrometry employs stable isotope-labeled analogs as internal standards (IS). Isotope-labeled analogs exhibit virtually identical chemical behavior to the 'native' substances, but are readily differentiated by their mass spectra. ID quantifi-
I
Phenylglyoxylic Acid (PGA)
Mandelic Acid (MA)
Hippuric Acid (HA)
t Author to whom correspondence should be addressed. 1052-93M/91/09051S7 $05.00
0 1991 by John Wiley & Sons, Ltd.
Figure 1. Structures of styrene metabolites.
Received 8 March 1991 Revised 13 May 1991
516
F. R. BROWN A N D W. M. DRAPER
cation can compensate for analytical error caused by incomplete extraction, electronic drift, changes in mass spectrometer detector response, and other variables. The use of ID mass spectrometry in clinical chemistry and pharmaceutical analysis is well established' 3,14 and ID also has become common in environmental and regulatory monitoring.' ,*16 In contrast, there have been few reports of the use of ID in LC/MS,"-" and, to our knowledge, none involving PB LC/MS instruments. This paper describes the ID quantification of model compounds in PB LC/MS. For this purpose, we examined the quantification of styrene's major mammalian metabolites, phenylglyoxylic (PGA), mandelic (MA) and hippuric (HA) acids (Fig. 1). A further objective was to evaluate coeluting, isotope-labeled analogs as carriers which might improve the transmission efficiency, linearity and sensitivity of PB LC/MS. Such improvements would greatly enhance the utility of PB LC/MS in occupational exposure monitoring, environmental analysis and other areas of regulatory analytical chemistry. EXPERIMENTAL Chemicals and reagents Acetonitrile (glass-distilled grade) and glacial acetic acid were obtained from E. M. Science (Cherry Hill, New Jersey, USA). Ammonium acetate was obtained from Sigma Chemical Co. (St Louis, Missouri, USA). PGA, MA, HA and deuterated styrene (d,, 98+ atom% enrichment) were obtained from Aldrich Chemical Co. (Milwaukee, Wisconsin, USA). d,-Ring-labeled benzaldehyde (98 atom% enrichment) and d,-ring-labeled benzoyl chloride (99 atom% enrichment) were obtained from Cambridge Isotope Laboratories (Woburn, Massachusetts, USA). Distilled, deionized water was used to prepare aqueous solutions. d,-Ring-labeled analogs of the styrene metabolites were prepared in our laboratory using established methods. d,-HA was obtained by reacting d,-benzoyl chloride with glycine under basic conditions." d,-PGA was prepared by oxidation of d,-styrene with alkaline potassium permanganate.22 d,-MA was prepared by the reaction of d,-benzaldehyde with dichlorocarbene (chloroform/ OH -) followed by hydrolysis in the presence of a phasetransfer ~atalyst.'~All products were purified by recrystallization and characterized by melting point, fast atom bombardment (FAB) mass spectrometry, proton nuclear magnetic resonance ('H-NMR), and EI and CI particle beam mass spectrometry. Experimental apparatus The LC/MS instrument consisted of an Isco Model 2300 HPLC pump, Valco 6-port injection valve with 10 p1 sample loop, a Hewlett-Packard 59980A PB interface, a Hewlett-Packard 5988A quadrupole mass spectrometer, and a Hewlett-Packard lo00 computer data system running RTE-A software. The system was tuned manually using perfluorotributylamine for both EI and CI modes. System performance was monitored by daily
injection of caffeine, and caffeine response was also used to adjust the nebulizer in accordance with the manufacturer's recommendations. Liquid chromatographylmassspectrometer operation For LC/MS determinations, a 3 pm, 60 x 4.6 mm C,, column (Hewlett-Packard) with a 0.5 pm filter and CIS guard column was used. The mobile phase was ammonium acetate buffer-acetonitrile (20 : 1, v/v). The buffer was prepared by adjusting 27 mM ammonium acetate to pH 3.1 with acetic acid. At higher flow rates, mobile phase built up on the tip of the nebulizer, adversely affecting instrument performance. Thus, a flow rate of 0.3 ml min-' was used, which is typical for PB LC/MS operation. The mass spectrometer was operated as follows: EI mode; electron multiplier, 1840 V; filament current, 300 PA; source pressure, 3 x lo-, torr. The unlabeled styrene metabolites were detected by selected ion moniPGA, MA, toring (SIM) of masses m/z 77 ([C,H,]'; HA), m/z 105 ([C6H5CO]+: PGA, HA) and m/z 107 ([C6H,CHOH)+: MA), and the labeled styrene metabolites were detected by monitoring masses of m/z 82 ([C,D,]+: d,-PGA, d,-MA, d,-HA), m/z 110 ([C,D,CO]+: d,-PGA, d,-HA) and m/z 112 ([C,D,CHOH]+: d,-MA), with a dwell time of 300 ms for each mass.
Experimental procedure Calibration studies. The effect of the use of internal standards on PB LC/MS calibration curves was studied as follows: four calibration curves were generated in which each set of standards for a given calibration curve was spiked with a constant level of internal standard. The levels of internal standard in the four calibration curves were 0, 50, 100 and 250 ng pl-' for each internal standard. The concentration of native compound for each set of standards in a given calibration curve was 0, 20, 50, 100, 250, 500 and lo00 ng pl-'. In an additional experiment, a calibration curve was generated in which the level of native compound ranged from 50 to 2000 ng pl-' with a level on internal standard of 500 ng pl'-' Determination of styrene metabolites in urine. To further evaluate isotope dilution PB LC/MS, urine samples from styrene-exposed subjects were analyzed. Urine specimens were obtained from investigators studying styrene toxicokinetics in human subjects. Styrene exposure has been associated with use of Styrofoam cups, plastics and smoking, and these exposures as well as dietary intake were controlled during the study. Adult male subjects inhaled styrene vapor in an exposure chamber at the San Francisco General Hospital, Lung Biology Center. The subjects were exposed on two days for three 100 min periods each day. On day 1, the subjects were exposed to 15, 32.5 and 50 ppm of styrene vapor, and on day 2 were exposed to 50,75 and 99 ppm styrene vapor.24 Urine samples were collected during the exposure, and 24 h urine samples were collected for three days following the final exposure.
QUANTITATION BY PARTICLE BEAM LC-MS
Urine samples were extracted with organic solvent prior to analy~is.~'Briefly, 1.0 ml of urine was combined with 0.10 ml of concentrated hydrochloric acid, 0.30 g of sodium chloride and 50 pl of a solution containing 10 pg pl-' each of d5-PGA, dS-MA and d5-HA, before extraction with 4.0 ml of diethyl ether-methanol (9 :1, v/v). Samples were stirred for 1.5 min before removing 2.0 ml of the organic layer, which was reduced to dryness under a stream of nitrogen and redissolved in 0.50 ml of mobile phase.
517
60
PGA
7
RESULTS AND DISCUSSION
Linearity of PB LC/MS Aerosol generation, the introduction of a dispersion gas to prevent aerosol agglomeration and reduce droplet size, desolvation at or near atmospheric pressure, and momentum separation all are fundamental processes of the PB liquid chromatograph/mass spectrometer interface.6 In the momentum separator the aerosol gas jet undergoes supersonic expansion-enrichment is achieved by skimming the expansion jet where components with the least momentum, e.g. gases and smalldiameter particles, are preferentially excluded. Analyte transport efficiencies are generally less than 10% for most analytes in the PB interface.s Analyte loss occurs primarily in the momentum separator due to particle sedimentation, misalignment of nozzles and skimmer cones, and turbulence.6 The mean diameter of desolvated particles is a function of analyte concentration6 and the diameter of the nebulizer tip.5 The small particles resulting at low analyte concentrations are more subject to turbulent losses, and thus have reduced transport efficiency. In the current study, calibration curves were generated under realistic LC/MS operating conditions with a reversed-phase column and 0.30 ml min-' of an ammonium acetate buffer-acetonitrile mobile phase. Calibration curves for the styrene metabolites (Fig. 2) clearly demonstrate the non-linearity of PB LC/MS
Figure 2. Calibration curves for model compounds. = M A ; A = HA.
= PGA;
response, even in the presence of ammonium acetate. The level of ammonium acetate used was well above that level previously reported to enhance linearity.l 2 As in previous reports the detector response was second order (quadratic). A typical chromatogram (full scan, total ion current) of a mixture of the three native compounds is shown in Fig. 3. Figure 4 shows the EI mass spectra obtained for the three native compounds in a single LC/MS run. Molecular ions are seen in each mass spectrum, but the m/z 107 major ions are m/z 105 ([C,H,O]'), m/z 77 ([c6H,]') and m/Z 79 ([c6H,]'). ([C,H,O]', Using the mass spectral library (NIST) and the mass spectrometer data system, the instrument was able to tentatively identify MA and HA. PGA was not in the library. The primary ions used for quantification are m/z 105 (PGA, H A ) and m/z 107 (MA). As discussed above, addition of mobile-phase modifiers such as ammonium acetateI2 or malic acid" has been reported to linearize the PB LC/MS response, presumably by contributing to the formation of heavier, high-momentum particles to which the analyte molecules are sorbed. As such, these mobile-phase modifiers
MA
HA
Figure 3. El total ion current chromatogram of non-deuterated model compounds. Mass spectrometric conditions: scan mode, m/z 65-300 with 1.96 s scan time; filament current = 300 pA, Liquid chromatographic conditions: 0.3 ml min-' mobile phase; 4.6 mm x 60 mm C,, column, 3 pm packing.
F. R. BROWN A N D W. M. DRAPER
518
The matrix effect in PB LC/MS
Phenylglyoxylic Acid 1200
200
70
8009 700d
600q
100
80
90 100 110 120 130 140 150 160 170 180 190
Mandelic Acid (107
tIio5
I
1
13
The Matrix Effect in Particle Beam Liquid Chromatography/MassSpectrometry and Reliable Quantification by Isotope Dilution F. Reber Brown and William M. Draper? Hazardous Materials Laboratory, 2151 Berkeley Way, Berkeley, California 94704, USA
The transport efficiency of the padcle beam liquid chromatography/mnss spectrometer interface is idueaeed by a d y t e concentration contributing to a widely reported non-linearity. In this work, coeluting, isotopelabeled interMIstandards were investigated as 'carriers' to improve the transport efficiency and linearity. Three styrene metabolitemadelk, phenyldyoxylic and bippuric acids-and tbei pentadeutero analogs were separated by reversed-phase liquid chromatogrophy (LC) with an ammonium acetate-acetonitrile mobile phase. Selected positive ions produd by electron ionization were monitored to generate pfllicle beam LC/MS calibration curves. The present study demonstrates that particle beam LC/MS not only is ma-linear, but also is subject to a matrix e&ect presumably by the same mechanism responsible for non-linearity. Caeluting, isotopelabeled internal standards were ineffective at linearizing the pnrticle beam liquid chromatograph/mass spectrometer detector response. Isotope dilution quantification, however, compensates for variable transport efficiencies, linearizes calibration and compensates for the matrix e&ect, affording reliable quantification of the styrene metabolites.
INTRODUCTION Particle beam liquid chromatography/mass spectrometry (PB LC/MS) has become popular because of ease of operation, compatibility with many LC mobile phases and the ability to operate in both electron (EI) and chemical ionization (CI) modes. Generation of library-searchable, EI spectra by PB LC/MS is an important advantage over other LC/MS techniques, and this capability is critical in regulatory monitoring and identification of unknowns. By comparison, direct liquid introduction LC/MS generates only CI spectra,'-' and thermospray LC/MS provides principally molecular weight inf~rmation.~.' Finally, because PB LC/MS uses a conventional mass spectrometric source, particle beam instruments are compatible with other mass spectrometric techniques such as GC/MS and sample introduction by solid probe. The development of PB LC/MS was first described by Willoughby and Browner in 19844 and the technique has been r e v i e ~ e d . ~As . ~ .PB ~ LC/MS is less than ten years old, and commercial PB interfaces have only been available for a few years, many of the basic features of the technique are only now being investigated. Recent reports have described characteristics of tuning, mobile phase effects, signal optimization and so forth.'-' Quantitative PB LC/MS has been hampered by nonlinear detector response. PB LC/MS non-linearity has been observed in various laboratoriesa*'0-'2 although at this time it has not been studied thoroughly or sys-
tematically. PB LC/MS calibration curves have been reported to be quadratic or second-order,a*'Obased on regression analysis of limited data sets. Trial and error and serendipity have led to means for minimizing and compensating for PB LC/MS non-linearity. Kim et al. added malic acid to the mobile phase to aid in linearizing the PB LC/MS response of daminozide." Similarly, Bellar and coworkers observed that ammonium acetate as a mobile phase modifier acted as a 'carrier' and extended the PB LC/MS linear dynamic range.I2 Isotope dilution (ID) mass spectrometry employs stable isotope-labeled analogs as internal standards (IS). Isotope-labeled analogs exhibit virtually identical chemical behavior to the 'native' substances, but are readily differentiated by their mass spectra. ID quantifi-
I
Phenylglyoxylic Acid (PGA)
Mandelic Acid (MA)
Hippuric Acid (HA)
t Author to whom correspondence should be addressed. 1052-93M/91/09051S7 $05.00
0 1991 by John Wiley & Sons, Ltd.
Figure 1. Structures of styrene metabolites.
Received 8 March 1991 Revised 13 May 1991
516
F. R. BROWN A N D W. M. DRAPER
cation can compensate for analytical error caused by incomplete extraction, electronic drift, changes in mass spectrometer detector response, and other variables. The use of ID mass spectrometry in clinical chemistry and pharmaceutical analysis is well established' 3,14 and ID also has become common in environmental and regulatory monitoring.' ,*16 In contrast, there have been few reports of the use of ID in LC/MS,"-" and, to our knowledge, none involving PB LC/MS instruments. This paper describes the ID quantification of model compounds in PB LC/MS. For this purpose, we examined the quantification of styrene's major mammalian metabolites, phenylglyoxylic (PGA), mandelic (MA) and hippuric (HA) acids (Fig. 1). A further objective was to evaluate coeluting, isotope-labeled analogs as carriers which might improve the transmission efficiency, linearity and sensitivity of PB LC/MS. Such improvements would greatly enhance the utility of PB LC/MS in occupational exposure monitoring, environmental analysis and other areas of regulatory analytical chemistry. EXPERIMENTAL Chemicals and reagents Acetonitrile (glass-distilled grade) and glacial acetic acid were obtained from E. M. Science (Cherry Hill, New Jersey, USA). Ammonium acetate was obtained from Sigma Chemical Co. (St Louis, Missouri, USA). PGA, MA, HA and deuterated styrene (d,, 98+ atom% enrichment) were obtained from Aldrich Chemical Co. (Milwaukee, Wisconsin, USA). d,-Ring-labeled benzaldehyde (98 atom% enrichment) and d,-ring-labeled benzoyl chloride (99 atom% enrichment) were obtained from Cambridge Isotope Laboratories (Woburn, Massachusetts, USA). Distilled, deionized water was used to prepare aqueous solutions. d,-Ring-labeled analogs of the styrene metabolites were prepared in our laboratory using established methods. d,-HA was obtained by reacting d,-benzoyl chloride with glycine under basic conditions." d,-PGA was prepared by oxidation of d,-styrene with alkaline potassium permanganate.22 d,-MA was prepared by the reaction of d,-benzaldehyde with dichlorocarbene (chloroform/ OH -) followed by hydrolysis in the presence of a phasetransfer ~atalyst.'~All products were purified by recrystallization and characterized by melting point, fast atom bombardment (FAB) mass spectrometry, proton nuclear magnetic resonance ('H-NMR), and EI and CI particle beam mass spectrometry. Experimental apparatus The LC/MS instrument consisted of an Isco Model 2300 HPLC pump, Valco 6-port injection valve with 10 p1 sample loop, a Hewlett-Packard 59980A PB interface, a Hewlett-Packard 5988A quadrupole mass spectrometer, and a Hewlett-Packard lo00 computer data system running RTE-A software. The system was tuned manually using perfluorotributylamine for both EI and CI modes. System performance was monitored by daily
injection of caffeine, and caffeine response was also used to adjust the nebulizer in accordance with the manufacturer's recommendations. Liquid chromatographylmassspectrometer operation For LC/MS determinations, a 3 pm, 60 x 4.6 mm C,, column (Hewlett-Packard) with a 0.5 pm filter and CIS guard column was used. The mobile phase was ammonium acetate buffer-acetonitrile (20 : 1, v/v). The buffer was prepared by adjusting 27 mM ammonium acetate to pH 3.1 with acetic acid. At higher flow rates, mobile phase built up on the tip of the nebulizer, adversely affecting instrument performance. Thus, a flow rate of 0.3 ml min-' was used, which is typical for PB LC/MS operation. The mass spectrometer was operated as follows: EI mode; electron multiplier, 1840 V; filament current, 300 PA; source pressure, 3 x lo-, torr. The unlabeled styrene metabolites were detected by selected ion moniPGA, MA, toring (SIM) of masses m/z 77 ([C,H,]'; HA), m/z 105 ([C6H5CO]+: PGA, HA) and m/z 107 ([C6H,CHOH)+: MA), and the labeled styrene metabolites were detected by monitoring masses of m/z 82 ([C,D,]+: d,-PGA, d,-MA, d,-HA), m/z 110 ([C,D,CO]+: d,-PGA, d,-HA) and m/z 112 ([C,D,CHOH]+: d,-MA), with a dwell time of 300 ms for each mass.
Experimental procedure Calibration studies. The effect of the use of internal standards on PB LC/MS calibration curves was studied as follows: four calibration curves were generated in which each set of standards for a given calibration curve was spiked with a constant level of internal standard. The levels of internal standard in the four calibration curves were 0, 50, 100 and 250 ng pl-' for each internal standard. The concentration of native compound for each set of standards in a given calibration curve was 0, 20, 50, 100, 250, 500 and lo00 ng pl-'. In an additional experiment, a calibration curve was generated in which the level of native compound ranged from 50 to 2000 ng pl-' with a level on internal standard of 500 ng pl'-' Determination of styrene metabolites in urine. To further evaluate isotope dilution PB LC/MS, urine samples from styrene-exposed subjects were analyzed. Urine specimens were obtained from investigators studying styrene toxicokinetics in human subjects. Styrene exposure has been associated with use of Styrofoam cups, plastics and smoking, and these exposures as well as dietary intake were controlled during the study. Adult male subjects inhaled styrene vapor in an exposure chamber at the San Francisco General Hospital, Lung Biology Center. The subjects were exposed on two days for three 100 min periods each day. On day 1, the subjects were exposed to 15, 32.5 and 50 ppm of styrene vapor, and on day 2 were exposed to 50,75 and 99 ppm styrene vapor.24 Urine samples were collected during the exposure, and 24 h urine samples were collected for three days following the final exposure.
QUANTITATION BY PARTICLE BEAM LC-MS
Urine samples were extracted with organic solvent prior to analy~is.~'Briefly, 1.0 ml of urine was combined with 0.10 ml of concentrated hydrochloric acid, 0.30 g of sodium chloride and 50 pl of a solution containing 10 pg pl-' each of d5-PGA, dS-MA and d5-HA, before extraction with 4.0 ml of diethyl ether-methanol (9 :1, v/v). Samples were stirred for 1.5 min before removing 2.0 ml of the organic layer, which was reduced to dryness under a stream of nitrogen and redissolved in 0.50 ml of mobile phase.
517
60
PGA
7
RESULTS AND DISCUSSION
Linearity of PB LC/MS Aerosol generation, the introduction of a dispersion gas to prevent aerosol agglomeration and reduce droplet size, desolvation at or near atmospheric pressure, and momentum separation all are fundamental processes of the PB liquid chromatograph/mass spectrometer interface.6 In the momentum separator the aerosol gas jet undergoes supersonic expansion-enrichment is achieved by skimming the expansion jet where components with the least momentum, e.g. gases and smalldiameter particles, are preferentially excluded. Analyte transport efficiencies are generally less than 10% for most analytes in the PB interface.s Analyte loss occurs primarily in the momentum separator due to particle sedimentation, misalignment of nozzles and skimmer cones, and turbulence.6 The mean diameter of desolvated particles is a function of analyte concentration6 and the diameter of the nebulizer tip.5 The small particles resulting at low analyte concentrations are more subject to turbulent losses, and thus have reduced transport efficiency. In the current study, calibration curves were generated under realistic LC/MS operating conditions with a reversed-phase column and 0.30 ml min-' of an ammonium acetate buffer-acetonitrile mobile phase. Calibration curves for the styrene metabolites (Fig. 2) clearly demonstrate the non-linearity of PB LC/MS
Figure 2. Calibration curves for model compounds. = M A ; A = HA.
= PGA;
response, even in the presence of ammonium acetate. The level of ammonium acetate used was well above that level previously reported to enhance linearity.l 2 As in previous reports the detector response was second order (quadratic). A typical chromatogram (full scan, total ion current) of a mixture of the three native compounds is shown in Fig. 3. Figure 4 shows the EI mass spectra obtained for the three native compounds in a single LC/MS run. Molecular ions are seen in each mass spectrum, but the m/z 107 major ions are m/z 105 ([C,H,O]'), m/z 77 ([c6H,]') and m/Z 79 ([c6H,]'). ([C,H,O]', Using the mass spectral library (NIST) and the mass spectrometer data system, the instrument was able to tentatively identify MA and HA. PGA was not in the library. The primary ions used for quantification are m/z 105 (PGA, H A ) and m/z 107 (MA). As discussed above, addition of mobile-phase modifiers such as ammonium acetateI2 or malic acid" has been reported to linearize the PB LC/MS response, presumably by contributing to the formation of heavier, high-momentum particles to which the analyte molecules are sorbed. As such, these mobile-phase modifiers
MA
HA
Figure 3. El total ion current chromatogram of non-deuterated model compounds. Mass spectrometric conditions: scan mode, m/z 65-300 with 1.96 s scan time; filament current = 300 pA, Liquid chromatographic conditions: 0.3 ml min-' mobile phase; 4.6 mm x 60 mm C,, column, 3 pm packing.
F. R. BROWN A N D W. M. DRAPER
518
The matrix effect in PB LC/MS
Phenylglyoxylic Acid 1200
200
70
8009 700d
600q
100
80
90 100 110 120 130 140 150 160 170 180 190
Mandelic Acid (107
tIio5
I
1
13
